VMFS Recovery

Recovers data from ESX server

VMFS Recovery™ scan cache guide and usage


Modern corporate requirements set very high limits for disk capacity. It's already quite usual to see datastores and iSCSI storages over 20Tb. However, the disk read speed parameter doesn't improve corresponding to capacity growth. Adding to the picture common data recovery issues such as bad blocks or RAID corruption cuts disk read speed even more. This makes downtime because of recovery bitter or even unacceptable.

Situation may become even worse if there are a power down or software crashes. Which requires starting the scan from scratch.

To deal with such situations Diskinternals have developed and implemented scan cache which saves pointers to file system structures and (if used during a repeating scan) allows reading from the scanned disk file system structures only. Using scan cache will not make the process instant, as file system structures are to be read anyway. Speed up depends on the amount of file system information to be processed but you may expect to get a reduction from a couple of days to 20 minutes.

Now to the details: there are two methods to set up a scan cache. I'd recommend setting up it before a scan is started, especially if you expect any issues such as power-downs or hardware issues.

Setting up cache before scan

Please click "View->Preferences" to open VMFS Recovery™ options (keyboard short key F11) and press CTRL+E to open Developer tab.

Check "Developer mode" and specify "Write sectors" and "Write Tree" with paths to some existing folder and add a file name. Just make sure it won't write anything to the damaged disk.

Please be advised that scan cache is set independently for each file system. So if you need to scan VMFS disk, you need to set up scan cache in VMFS section.

Developer options

Normally we place some temporary path at the system disk like: "C:\diskinternals\sectors.cache" and "C:\diskinternals\tree.cache" or "C:\temp\sectors" and "C:\temp\tree". We use C: disk because it's normally a Windows system disk. Actual file names, extension, and folder can be changed to any that suits you.

Sectors and Tree parameters

Sectors parameter is a scan cache. It contains only pointers to the disk and doesn't have even a single bit of user information.

“Tree” parameter saves file and folder tree which is required by our developers and engineers to check the status of your case. I'd advise creating this log file just in case it would improve the investigation of your case if you would need our support. Creating a “Tree” log will not affect the performance of scan speed. Data saved in the “Tree” log contains file and folder names plus contents of small files, as contents of very small files (less than 256kb) are stored within file system contents without real disk space allocation.

Approximate cache size

Scan cache size depends on the amount of file system information on the disk. The larger disk you are recovering data from and the more files and data it has, the larger the scan cache would be. Normally it is to be expected to be between 1-10Gb.

Saving scan cache after scan is complete

To do this you need to enable Developer mode by clicking "View->Preferences" to open options and press CTRL+E to open Developer tab. Check "Developer mode". Now return to the disk list and right-click on the previously scanned disk and select the "Save Cache" option.

saving scan cache from context menu

Please note: should you expect any problems, it would be much better to set up a scan before scan rather than hoping it to be finished first.

Additional logging and options

It is possible to set up additional logging about software work at the Developers page. It may be useful if you experience any software crashes. To do this specify a path with file name at the “File Name Log” parameter, check “Additional debug msg” and “Save Log” options.

Setting up additional scan logging and debug messages

You may also find useful a “Flush” option at the sections corresponding to processed file systems. It will write any new information immediately instead of storing it in the buffer. This reduces scan performance but valuable for bad block recognition or if there is some software crash to be fixed.

Using previously stored scan cache

In order to use previously saved scan cache, please reopen Developer tab as described above and specify path to scan cache at the "Read Sectors" parameter in corresponding section. Change “Write Sectors” parameters to use different file name.

If everything was set up correctly, you'll see a following warning message: “Sector cache found with sectors inside. Would you like to use it?”

Cache usage warning

Click “Yes” to process scan cache.

Cleaning the settings

Please be advised: scan cache will be used every time disk with the corresponding file system is opened. Currently, there is no check for different disks and thus once set up scan cache for VMFS(or any other) disk will be applied for any other VMFS disk, which will lead to incorrect results of the scan. To avoid such a situation it's important to clear all setting to default, I'd advise deleting the “FSM.ini” file at the software's directory.

Return to contents